Reader Brian Thomas forwarded us the following very detailed letter describing his experience as a recount worker.

Click extended entry to read the whole thing. Such confidence of a "no vote" should mean a "vote of no confidence," in this election.

December 24, 2004

Dino Rossi Campaign 15100 SE 38th Street #715 Bellevue, Washington 98006

Re: A ‘No Vote’ Ballot Counted As A Gregoire Vote

On Monday, December 20, 2004, in a second count of a box of absentee ballots I found, in the middle of ballots with Christine Gregoire votes, a ballot on which the bubble next to Christine Gregoire had been filled in and ‘X’ed out twice. ‘X’ing out a bubble being the correct procedure for canceling a vote on an absentee ballot. Next to the ‘X’ed out bubble was printed the words “no vote” with an arrow pointing to the name Christine Gregoire. I showed the ballot to the recorder and the Democrat counter and held it so that it could be seen by the observer sitting next to me. I handed the ballot to the recorder to be sent to the canvas board. When filling out the tally slips the ballot was again viewed and noted by the Democrat counter and myself as a canvas ballot and by the recorder on the recorder’s tally slip as a canvas ballot. The ballot was handed to the runner as a canvas ballot with the box of counted ballots

“How and why did the ballot end up in the previously counted Gregoire ballots?

Our count was the second count of that precinct as there was a previous tally slip in the sealed box when it was opened. I was at least the third person to have viewed that ballot as it would had been viewed by the two counters during the previous count. Yet a ballot clearly marked as to not be a vote for Christine Gregoire was in with the ballots with Gregoire votes and not separated out in a way that would denote that it had been counted as anything other than a Gregoire vote.

When the ballots from a precinct are hand counted, prior to counting they were grouped into six (6) separate stacks: Rossi, Gregoire, Bennett, write-in, under-vote and over-vote / canvas board. At the completion of the count the ballots are usually placed back in the box in these groupings or as a single stack containing these groupings as it easier to handle the ballots as stacks and there is no reason to shuffle them. As an example, the stack of Rossi ballots would be placed in the bottom of the box with the Gregoire, Bennett, etc. placed on top until all the ballots are placed in the box or the groups are placed in a single stack on the table and the single stack is placed in the box.

When the ballots are taken out of the box for a recount, the box is usually turned over and the intact stack of ballots dumped on the table. Some people take the ballots out of the box in handfuls and place them in a stack on the table in the order that they are taken out of the box. In neither method, as both a counter and an observer, did I ever see the ballots shuffled and the groups of ballots disrupted except when the ballots were taken out in several handfuls and the groups of ballots arbitrarily broken in the process. In the latter situation, a non conforming ballot(s) could be placed within a stack of the same grouping if the ballots were removed from the box by the handful and, for example, the first handful was of a portion of the Rossi ballots and the second handful was of the remaining Rossi ballots and a ballot(s) from the next grouping, such as a write-in ballot(s). When the second handful is placed on top of the first handful, there could be a non conforming ballot(s), such as the write-in ballot(s), in the stack of Rossi ballots.

On the day that I found the ‘no vote’ ballot, at our table I was the one who opened the sealed box of ballots and took the ballots out of the box. My procedure was if I could reach into the box and take the ballots out in one handful I would. Otherwise I would turn the box on it’s side, holding the ballots in, and then turn the box upside down, depositing the ballots in a stack on table top. If I was to take the ballots out of the box in more than one handful, there would have been an arbitrarily break line but at no time was there a shuffling of ballots. In the box of ballots that contained the ‘no vote’ ballot I dumped the ballots in one stack on the table.

Either way, the ‘no vote’ ballot should not have been in the Gregoire ballots, whether the ballots were dumped out in a stack or were taken out of the box in a series of handfuls and placed in a stack. In the first instance, the ‘no vote’ ballot should not have been in the Gregoire stack as it was clearly not a Gregoire vote. In the second instance, that being the ‘no vote’ ballot being in one of the other groupings and having been arbitrarily picked up as part of a hand full and placed in the stack in the process of restacking, it should not have been in the Gregoire stack. The ballot should not have been in the box of ballots in the first place. It did not meet the requirements of any of the required groupings and should have sent to canvas for disposition at the time of the first hand count.

Why was the ‘no vote’ ballot in the box and why was the vote on the ballot being counted?

Why the ballot was being counted as a Gregoire vote is at this point speculation, but it is not without reason and the reason for that ballot to be in the box of ballots after the first hand count is open to question and examination.

The ballots in the box were ballots that had been previously hand counted. The box was sealed and would have arrived with any additions or deletions to the box of ballots having been concluded with the previous hand count. The only source of an additional ballot(s) would be a ballot(s) from a hand count of a different precinct that had been removed from that hand count as being in the wrong precinct. That ballot(s) would have been delivered in a separate envelope with the box of ballots. This was not the case. And if a ballot(s) had been delivered in a separate envelope with the box of ballots it would not have been among the Gregoire ballots as the procedure was for me to open the box of ballots and stack them on the table and for the other counter to open the envelope and to place the ballot(s) on top of the stack of the ballots from the box. The ‘no vote’ ballot was in the group of ballots with Gregoire votes in the sealed box of ballots.

During the hand count each ballot is viewed at least three times. Each table had it’s own procedure. Our procedure was to separate the ballots into two piles, three if the recorder was participating, and each ballot was checked for the correct precinct number. The stacks of ballots were then exchanged and if the recorder was participating in the precinct check the recorders stack was separated into two stacks and which were placed on the two counters stacks. Each counter then checked the other counters stack plus half of the recorders stack. We then each separated our stacks of ballots into the six (6) separate stacks: Rossi, Gregoire, Bennett, write-in, under-vote and over-vote / canvas board with the recorder observing the separation. Some counters worked off of the same stack of ballots at the same time with each counter removing and stacking ballots for a particular candidate, such as the Republican removing the ballot with the Rossi vote and the Democrat removing the ballot with the Gregoire vote. If the recorder was participating, the recorder would remove and stack the ballots for Bennett, write-in, under-vote and over-vote / canvas board.

After the ballots were separated into their distinct stacks, we each checked the stacks to confirm that the stack only contained ballots with votes for that individual, Rossi, Gregoire or Bennett, or grouping, write-in, under-vote and over-vote / canvas board. (It was during this procedure that I pulled the ‘no vote’ ballot for the canvas board.) We would then separately count the votes in the stack of Rossi or Gregoire ballots, exchanging the stacks and individually writing down our count. We would then separately count the ballots with Bennett votes and the ballots of the other groupings, write-in, under-vote and over-vote / canvas board, individually writing down our count. Some counters counted the votes as they checked the stacks to confirm that the stack only contained ballots particular to that grouping. We did not use that method as if upon exchanging the stacks of ballots a ballot was questioned, such as a ballot going to the canvas board, the stack would have to be recounted by the other counter. At completion of the counts the individual tally sheets were handed to the recorder to compare the tallies. If the tallies were a match, the ballots were placed back in the box. Since after the ballots had been counted they were in their individual stacks, the individual stacks were placed in the box or the individual stacks were stacked upon each other and the combining stack was placed in the box. The only single ballots would have been from a grouping that contained only one ballot. The stacks were not shuffled.

During the procedure that we used at our counting table, if the ‘no vote’ ballot had not been pulled it would have been viewed by two different people three times for total of being viewed five (5) times. Once when the ballot was separated from the stack as a Gregoire vote, twice when the Gregoire vote stack was checked by each counter to confirm that it only contained Gregoire votes and twice when the Gregoire stack was counted by each counter. At our counting table the recorder observed each of these steps for an additional three times that the ‘no vote’ ballot would have been viewed. I pulled the ‘no vote’ ballot at the first stage of our ballot counting procedure, separating the ballots, as it should have been done by the previous counters. The question is; why wasn’t it?

Why was the ‘no vote’ ballot in the box?

Our vote counting procedure included checks at each stage of the procedure. Counting procedures may vary but the question as to how and why the ‘no vote’ ballot had been counted as an Gregoire vote three times can not be simply dismissed as a difference in procedure. Regardless of the procedure, each ballot should have been viewed by both counters at least twice. If the stack of ballots were separated by the two counters simultaneously and then the individual stacks of ballots were each counted by each counter, each counter would have seen the ‘no vote’ ballot twice. Once at the separation of the ballot into the individual candidates stack and once when the votes were counted. Unless the ballots were counted and not the vote for the candidate, in which case a counter could have viewed the vote only once Which would bring up the competency of the counters and the King County Elections staff for allowing counters to count ballots and not requiring them to count ‘votes’.

If the counters counted the ballots and not the votes, they would have both seen the ballot and the ‘no vote’ at the time of separation of the ballots into the candidates individual stacks. To be able to separate the ballots into the individual candidates stack the counters have to look at the candidates to make the decision as to which candidate the vote belongs.. For the purpose of an excuse as to why the ‘no vote’ ballot was in the Gregoire stack at the start of the second hand recount, that being that sloppy examination and counting procedure of two people, one a Republican and one a Democrat, resulting in missing a canceled Gregoire vote and including the vote as a Gregoire vote, still brings up the question as to why the ‘no vote’ ballot was in the recount ballots in the first place.

The ballot would have gone through the initial machine count and the machine recount as a vote for Gregoire as the machine would have simply read the filled in bubble next to Gregoire’s name and would not have read the ‘X’ed out bubble as cancellation of the vote, nor the ‘no vote’ notation next to Gregoire’s name. Since after canceling the initial vote a subsequent vote was not made, if it had been the machine would have read it as an over-vote, rejected the ballot and it would have sent to the ‘enhancement’ table during the machine count, I asked King County Elections staff the procedure for absentee ballots in which a vote is canceled but a subsequent is not made, in effect a under-vote. I was told that each absentee ballot is checked for incongruities at the time of opening and at that time problem would be resoled. Evidently that check and subsequent resolution of the problem was not resolved as the vote was counted for Gregoire. The question is why.

Why was a canceled vote for Gregoire allowed to be counted as a vote for Gregoire? Thrice.

If, contrary to what I was told by the King County Elections staff, absentee ballots upon opening are not checked for irregularities that could result in an inaccurate vote count by the tabulation machines prior to being read by the tabulation machines, why aren’t ballots checked for problems that could, and in this case evidently did, result in the casting a false vote for Christine Gregoire and in an inaccurate vote tally? This is a simple problem and one that should have been foreseen by the King County Election staff. For King County Elections staff to have not foreseen this as a potential problem, one that could result in false readings by the tabulation machines and inaccurate vote tallies, would be gross negligence and incompetence and would be tantamount to misfeasance by King County Elections staff and should be investigated as such as it could be indicative of a larger problem of malfeasance by King County Elections supervisors.

If the absentee ballots upon opening are checked for irregularities that would result in an inaccurate vote count by the tabulation machines prior to being read by the tabulation machines, how and/or why was this ballot not ‘corrected’, as were ballots with other problems or the intent was not clear as the ‘no vote’ ballot, prior to being evidently counted as a false vote for Gregoire resulting in an inaccurate vote tally? If the failure of the King County Elections staff to take steps necessary to insure that the correct vote is tallied is simply the due to gross negligence and incompetence is one thing. But if the King County Elections staff knew that the ‘no vote’ ballot would be counted as a Gregoire ballot by the tabulation machines unless the ballot was rejected for some other reason and choose to not make the necessary corrections to prevent the vote being cast for Gregoire, not considering a possible hand count, that would be malfeasance by King County Elections staff and would be tantamount to intent to commit fraud and should be investigated.

Brian Thomas

Posted by Brian Crouch at January 02, 2005 04:52 PM | Email This